| In recent years,blockchain technology has developed vigorously.From the first single digital currency application to today's multiple application fields such as finance,energy,information security,education,urban governance,food and e-commerce,blockchain technology has been deeply studied and explored.The establishment of blockchain system is inseparable from the support of underlying technology,and the efficiency of consensus algorithm will affect the security and practicability of the whole system,which determines the application direction of blockchain.Therefore,designing an efficient consensus algorithm is of great significance to improve the performance of the blockchain system and guarantee the security of information.This paper studies the performance and application of blockchain consensus algorithm.Firstly,an improved PBFT algorithm based on dynamic classification is proposed;Secondly,for the alliance chain,an improved alliance chain consensus algorithm based on subgroup partition is proposed,which solves the problems of poor dynamics and low throughput;Finally,the popular applications of blockchain are studied,and the intelligent property management system based on Ethereum is designed and implemented.The main work of this paper is as follows:(1)An improved PBFT algorithm based on dynamic classification is proposed.First,the node behavior scoring model and node ranking model are designed in DLPBFT algorithm,which reduces the amount of nodes taking part in the consensus,improves the enthusiasm of nodes participating in the consensus,and guarantees the security of the system.Secondly,the selection method of primary node in traditional PBFT algorithm is changed.By selecting the primary node at the highest level,the amount of view switching is decreased and the performance of consensus is enhanced.Finally,experimental and simulation comparisons show that the improved DLPBFT algorithm has better performance in throughput,latency,security and dynamics.(2)An improved coalition chain consensus algorithm based on subgroup division is proposed.First,the SD-PBFT algorithm decomposes the number of consensus messages and improves the efficiency of consensus by getting the location coordinates and number of nodes in the federation chain and dividing all nodes in the network into subgroups.Secondly,the node with the highest score of node behavior in the subgroup is taken as the upper node to take turns as the primary node,which ensures the security of the primary node and reduces the number of view switching.Finally,experimental and simulation comparisons show that the improved SD-PBFT algorithm has better performance in throughput,latency,robustness and dynamic,and is more suitable for federation chain scenarios.(3)Directing against the extant problems of property management in smart city,an intelligent real estate management system based on Ethereum is designed and implemented by using the alliance chain,which solves the problems such as residents' distrust of property management,difficult data disclosure and easy tampering.First,the requirements of the intelligent property management system are analyzed in depth,and the system is designed in detail.Secondly,the block chain is used as the bottom support technology to achieve the de-centralization.Finally,the smart contract of Ethereum is used to realize the chain storage of system information,which guarantees the security,reliability and untampered modification of the real estate management system.Through the research in this paper,DLPBFT algorithm and SD-PBFT algorithm are proposed.The experimental results show that the algorithm in this paper gains better results in various aspects of consensus performance than traditional PBFT algorithm.It is meaningful to further study the theory and application of block chain. |
PDF Full Text Request